Predictive Molecular Thermodynamic Models for Liquid Solvents, Solid Salts, Polymers, and Ionic Liquids

Abstract

solubility, swelling degree, and crystallinity of polymers in polymer processing.Undoubtedly, predictive molecular thermodynamic models are very important in separation processes and polymer processing. In the separation processes (e.g., extractive distillation, liquid-liquid extraction, absorption, etc.), a third solvent is commonly needed to add into the components to be separated so as to improve the separation factor. 1,2 So the solvent (entrainer or separating agent) is the core, and a suitable solvent plays an important role in the economical design of separation processes. However, it is tiresome to choose the best solvent from thousands of different substances for a given system through experiments. We should identify the relation between molecular structure of the solvent and separation performance. In this case, predictive molecular thermodynamic models are used as a screening tool to find out the best suited solvent rapidly. Only on this

Group contribution lattice fluid equation of state for CO2–ionic liquid systems: An experimental and modeling study

“…4,16,36,44,45 Thus, the additional Debye-H€ uckel term accounting for long-range (LR) electrostatic contributions can be avoided.…”

“…Thus, for a better and thorough understanding on the separation and thermodynamic behaviors of CO 2 -IL systems, reliable and applicable predictive models are indispensable. [4][5][6][7] The predictive models for treating the phase equilibria of the systems with ILs include activity coefficient models, equations of state, molecular simulations, neural networks, and so on, among which activity coefficient models and equations of state are commonly used due to the solid thermodynamic foundation and fast calculation speed. However, the activity coefficient models, including regular solution theory (RST), 8 universal quasichemical functional-group activity coefficients (UNIFAC), 4,[9][10][11] and conductor-like screening model for real solvent (COSMO-RS) 12-15 models, are not functions of volume, and also not dependent on pressure.…”

“…[4][5][6][7] The predictive models for treating the phase equilibria of the systems with ILs include activity coefficient models, equations of state, molecular simulations, neural networks, and so on, among which activity coefficient models and equations of state are commonly used due to the solid thermodynamic foundation and fast calculation speed. However, the activity coefficient models, including regular solution theory (RST), 8 universal quasichemical functional-group activity coefficients (UNIFAC), 4,[9][10][11] and conductor-like screening model for real solvent (COSMO-RS) 12-15 models, are not functions of volume, and also not dependent on pressure. On the other hand, the equations of state, such as nonrandom lattice-fluid equation of state (NLF EOS), 16,17 the series of statistical associating fluid theory (SAFT) EOS [18][19][20][21][22][23][24][25][26][27][28] (e.g., perturbed-chain polar SAFT [PCP-SAFT], soft-SAFT, heterosegmented-SAFT, and SAFT-c), and groupcontribution equation of state (GC EOS), 29 would take into account the dependence of phase volume on pressure for calculating the phase equilibria, which is especially important for estimating volume expansivity.…”

Study of Ionic Liquids as Entrainers for the Separation of Methyl Acetate–Methanol and Ethyl Acetate–Ethanol Systems Using the COSMO-RS Model

“…25 These conventional models require the experimental ternary vapor liquid equilibrium data of azeotropic mixtures containing ionic liquids, to estimate the interaction parameters of the models. 26 The UNIFAC model can be used to predict the activity coefficient at infinite dilution for polar systems. 26 The deviation of UNIFAC predictions of relative volatility with experimental data was 9.85% (average relative deviation) for the ethanol− water system.…”

Vapour permeation modelling

“…They are indeed able to describe pure components properties as well as gaspolymers system which cannot be easily addressed with normal activity coefficient approaches (Lei, Chen, Li, & Liu, 2008).…”